Optimal object matching via convexification and composition

Hongsheng Li, Junzhou Huang, Shaoting Zhang, Sharon Xiaolei Huang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

12 Citations (Scopus)

Abstract

In this paper, we propose a novel object matching method to match an object to its instance in an input scene image, where both the object template and the input scene image are represented by groups of feature points. We relax each template point's discrete feature cost function to create a convex function that can be optimized efficiently. Such continuous and convex functions with different regularization terms are able to create different convex optimization models handling objects undergoing (i) global transformation, (ii) locally affine transformation, and (iii) articulated transformation. These models can better constrain each template point's transformation and therefore generate more robust matching results. Unlike traditional object or feature matching methods with "hard" node-to-node results, our proposed method allows template points to be transformed to any location in the image plane. Such a property makes our method robust to feature point occlusion or mis-detection. Our extensive experiments demonstrate the robustness and flexibility of our method.

Original languageEnglish (US)
Title of host publication2011 International Conference on Computer Vision, ICCV 2011
Pages33-40
Number of pages8
DOIs
StatePublished - Dec 1 2011
Event2011 IEEE International Conference on Computer Vision, ICCV 2011 - Barcelona, Spain
Duration: Nov 6 2011Nov 13 2011

Publication series

NameProceedings of the IEEE International Conference on Computer Vision

Conference

Conference2011 IEEE International Conference on Computer Vision, ICCV 2011
CountrySpain
CityBarcelona
Period11/6/1111/13/11

Fingerprint

Convex optimization
Chemical analysis
Cost functions
Experiments

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Vision and Pattern Recognition

Cite this

Li, H., Huang, J., Zhang, S., & Huang, S. X. (2011). Optimal object matching via convexification and composition. In 2011 International Conference on Computer Vision, ICCV 2011 (pp. 33-40). [6126222] (Proceedings of the IEEE International Conference on Computer Vision). https://doi.org/10.1109/ICCV.2011.6126222
Li, Hongsheng ; Huang, Junzhou ; Zhang, Shaoting ; Huang, Sharon Xiaolei. / Optimal object matching via convexification and composition. 2011 International Conference on Computer Vision, ICCV 2011. 2011. pp. 33-40 (Proceedings of the IEEE International Conference on Computer Vision).
@inproceedings{ed4ac60479ce4264ae1252d290c3f39a,
title = "Optimal object matching via convexification and composition",
abstract = "In this paper, we propose a novel object matching method to match an object to its instance in an input scene image, where both the object template and the input scene image are represented by groups of feature points. We relax each template point's discrete feature cost function to create a convex function that can be optimized efficiently. Such continuous and convex functions with different regularization terms are able to create different convex optimization models handling objects undergoing (i) global transformation, (ii) locally affine transformation, and (iii) articulated transformation. These models can better constrain each template point's transformation and therefore generate more robust matching results. Unlike traditional object or feature matching methods with {"}hard{"} node-to-node results, our proposed method allows template points to be transformed to any location in the image plane. Such a property makes our method robust to feature point occlusion or mis-detection. Our extensive experiments demonstrate the robustness and flexibility of our method.",
author = "Hongsheng Li and Junzhou Huang and Shaoting Zhang and Huang, {Sharon Xiaolei}",
year = "2011",
month = "12",
day = "1",
doi = "10.1109/ICCV.2011.6126222",
language = "English (US)",
isbn = "9781457711015",
series = "Proceedings of the IEEE International Conference on Computer Vision",
pages = "33--40",
booktitle = "2011 International Conference on Computer Vision, ICCV 2011",

}

Li, H, Huang, J, Zhang, S & Huang, SX 2011, Optimal object matching via convexification and composition. in 2011 International Conference on Computer Vision, ICCV 2011., 6126222, Proceedings of the IEEE International Conference on Computer Vision, pp. 33-40, 2011 IEEE International Conference on Computer Vision, ICCV 2011, Barcelona, Spain, 11/6/11. https://doi.org/10.1109/ICCV.2011.6126222

Optimal object matching via convexification and composition. / Li, Hongsheng; Huang, Junzhou; Zhang, Shaoting; Huang, Sharon Xiaolei.

2011 International Conference on Computer Vision, ICCV 2011. 2011. p. 33-40 6126222 (Proceedings of the IEEE International Conference on Computer Vision).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Optimal object matching via convexification and composition

AU - Li, Hongsheng

AU - Huang, Junzhou

AU - Zhang, Shaoting

AU - Huang, Sharon Xiaolei

PY - 2011/12/1

Y1 - 2011/12/1

N2 - In this paper, we propose a novel object matching method to match an object to its instance in an input scene image, where both the object template and the input scene image are represented by groups of feature points. We relax each template point's discrete feature cost function to create a convex function that can be optimized efficiently. Such continuous and convex functions with different regularization terms are able to create different convex optimization models handling objects undergoing (i) global transformation, (ii) locally affine transformation, and (iii) articulated transformation. These models can better constrain each template point's transformation and therefore generate more robust matching results. Unlike traditional object or feature matching methods with "hard" node-to-node results, our proposed method allows template points to be transformed to any location in the image plane. Such a property makes our method robust to feature point occlusion or mis-detection. Our extensive experiments demonstrate the robustness and flexibility of our method.

AB - In this paper, we propose a novel object matching method to match an object to its instance in an input scene image, where both the object template and the input scene image are represented by groups of feature points. We relax each template point's discrete feature cost function to create a convex function that can be optimized efficiently. Such continuous and convex functions with different regularization terms are able to create different convex optimization models handling objects undergoing (i) global transformation, (ii) locally affine transformation, and (iii) articulated transformation. These models can better constrain each template point's transformation and therefore generate more robust matching results. Unlike traditional object or feature matching methods with "hard" node-to-node results, our proposed method allows template points to be transformed to any location in the image plane. Such a property makes our method robust to feature point occlusion or mis-detection. Our extensive experiments demonstrate the robustness and flexibility of our method.

UR - http://www.scopus.com/inward/record.url?scp=84863064181&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84863064181&partnerID=8YFLogxK

U2 - 10.1109/ICCV.2011.6126222

DO - 10.1109/ICCV.2011.6126222

M3 - Conference contribution

SN - 9781457711015

T3 - Proceedings of the IEEE International Conference on Computer Vision

SP - 33

EP - 40

BT - 2011 International Conference on Computer Vision, ICCV 2011

ER -

Li H, Huang J, Zhang S, Huang SX. Optimal object matching via convexification and composition. In 2011 International Conference on Computer Vision, ICCV 2011. 2011. p. 33-40. 6126222. (Proceedings of the IEEE International Conference on Computer Vision). https://doi.org/10.1109/ICCV.2011.6126222